Can the best traders reliably exploit the ex-dividend effect (the tendency for dividend-paying stocks to fall by less than the dividend amount after paying the dividend)? In their March 2014 paper entitled “Ex-Dividend Profitability and Institutional Trading Skill”, Tyler Henry and Jennifer Koski examine whether highly skilled traders bearing very low transaction costs (some institutions) successfully exploit this effect. They use actual transaction prices and actual transaction costs. They segment their sample period into three regimes: Regime 1 is pre-decimalization and pre-tax reform that equalizes capital gains and dividend tax rates; Regime 2 is post-decimalization and pre-tax reform; and, Regime 3 is post-decimalization and post-tax reform. They specify the ex-dividend trading window as days -5 through +5 relative to ex-dividend day 0. Using a large proprietary set of institutional common stock buy and sell transactions and associated transaction costs, and contemporaneous dividends, returns, bid/ask prices and trading volumes for those stocks, during 1999 through 2007 (24,741 ex-dividend events for 1,351 firms), they find that:
- On average over the entire sample period, tracked institutions represent about 17% of trading volume, with average percentage bid-ask spread and commission 0.29% and 0.12%, respectively.
- Average bid-ask spread falls from 0.62% in Regime 1 to 0.31% in Regime 2, and falls again to 0.15% in Regime 3.
- Average trading commissions are 0.15% during Regimes 1 and 2, and 0.09% during Regime 3.
- Institutional volume rises during ex-dividend events, more for stocks with high yields and less for stocks with high idiosyncratic risk.
- Based on event window closing stock prices over the full sample period, value-weighted average raw (market-adjusted) gross ex-dividend effect return is 0.17% (0.10%). Average raw (market-adjusted) ex-dividend effect returns vary across Regimes 1 through 3 from 0.24% to 0.09% to 0.16% (0.20% to 0.10% to 0.06%).
- Over the entire sample period, average actual pre-commission (post-commission) ex-dividend effect institutional return is 0.11% (-0.06%) based on value weighting.
- Results are similar for equal weighting.
- Though transaction costs shrink in recent years, average market-adjusted ex-dividend effect return shrinks even more.
- However, institutions tend to focus on certain ex-dividend events, and they earn significantly higher profits around these events.
- Also, institutions that exhibit trade execution skill unrelated to ex-dividend capture successfully exploit the ex-dividend effect. A one standard deviation increase in trader skill corresponds to a 0.10% boost in post-commission ex-dividend effect profitability, such that the top tenth of traders outperform the bottom tenth by an average 0.40%.
In summary, evidence indicates that highly skilled, low-cost traders successfully exploit the ex-dividend effect for U.S. stocks.
Cautions regarding findings include:
- Other investors may not be able to achieve the low transaction costs borne by these institutions.
- Other investors may not be able to execute trades as proficiently as top institutional managers.
See also “NASDAQ vs. NYSE Dividend Capture”.
